WO2021152585A1 - Screening system for abiotic stress mitigation in plants - Google Patents

Screening system for abiotic stress mitigation in plants Download PDF

Info

Publication number
WO2021152585A1
WO2021152585A1 PCT/IL2021/050094 IL2021050094W WO2021152585A1 WO 2021152585 A1 WO2021152585 A1 WO 2021152585A1 IL 2021050094 W IL2021050094 W IL 2021050094W WO 2021152585 A1 WO2021152585 A1 WO 2021152585A1
Authority
WO
WIPO (PCT)
Prior art keywords
plant
plants
data sets
data
plant populations
Prior art date
Application number
PCT/IL2021/050094
Other languages
French (fr)
Inventor
Dotan BORENSTEIN
Ṛcā GODBOLÉ
Sharon Devir
Original Assignee
Salicrop
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Salicrop filed Critical Salicrop
Publication of WO2021152585A1 publication Critical patent/WO2021152585A1/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • A—HUMAN NECESSITIES
    • A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
    • A01H—NEW PLANTS OR NON-TRANSGENIC PROCESSES FOR OBTAINING THEM; PLANT REPRODUCTION BY TISSUE CULTURE TECHNIQUES
    • A01H1/00—Processes for modifying genotypes ; Plants characterised by associated natural traits
    • A01H1/04—Processes of selection involving genotypic or phenotypic markers; Methods of using phenotypic markers for selection

Definitions

  • the present disclosure relates to a system and methods for automated Al-based phenotypic screening of plants, particularly under abiotic stress, and predicting plant performance in the field.
  • US patent 9025831 discloses an apparatus for screening potted plants and an automated method for high-throughput phenotypic screening of a plurality of plants.
  • the method comprises growing a population of plants, which may comprise one or more transgenic events, in a controlled environment for a defined growing period and are subjected to at least one induced stress.
  • the growth profile comprises at least three measurable characteristics selected from the group consisting of canopy area, chlorophyll content, anthocyanin content, biomass, plant height, root mass, water content, yield, the amount of water applied during the growing period, and water use efficiency.
  • individual plants can be selected from the population of plants based upon the determined growth profile.
  • Patent application WO2014/124128 discloses systems and methods for plant stress mitigation.
  • the computer-implemented system which in first aspect, provides a self-standardizing algorithm that can be universally applied to detect stress-related changes in plants stress before it is visible to the naked eye and objectively quantify that stress over time and space.
  • US patent 10492374 discloses a method for acquiring data associated with a plant growing in a field using a sensor, analyzing the data obtained from the sensor to extract, while plants still grow in the field, one or more phenotypes, determining one or more plant insights based on phenotypic traits, including information about the plant’s health, yield, growth, harvest, management, performance, or state; and generating a plant insights report.
  • the analysis of the data obtained from the sensor includes: predicting the phenotypic traits based on the sensor data and a computerized model; displaying indications of the phenotypic traits predicted; and obtaining a confirmation, modification, or addition indication from the user for at least one of the indications of the phenotypic traits predicted based on direct observation of the plant in the field by the user.
  • US Patent application 2011/0125477A1 discloses methods and related devices for predicting the presence or level of one or more characteristics of a plant or plant population based on spectral, multi-spectral, or hyper- spectral data obtained by, e.g., remote sensing.
  • the predictions and estimates furnished by the inventive methods and devices are useful in crop management, crop strategy, and optimization of agricultural production.
  • the method of estimating a plant characteristic comprises: 1) building a predictive model using inverse modeling using: i. a first set of spectroscopic data from a first plant population, and ii. corresponding measured characteristic data sets from the first plant population; and, 2) applying the model to a second set of spectroscopic data from a second plant, a second plant population, or both, so as to estimate the characteristic in the second plant.
  • an artificial intelligence-based system capable of (a) screening populations of plants in laboratory/small-scale settings for phenotypic differences and traits, under various environmental conditions or treatments, (b) performing analyses based on the data gathered from the laboratory experiments, and (c) translating said data into practical actions and recommendations for growers and seed companies by a unique algorithm, which can predict how individual plants/genotypes will perform in the field overtime under specific conditions, including abiotic stresses.
  • the present invention discloses such a system and algorithm configured to monitor plant performance under abiotic stress in early stages of crops in a relatively short period of time, and at relatively small number of plants and generate corresponding predictive recommendations and indications for growers.
  • FIG.l depicting a schematic flowchart of the system of the present application
  • FIG.2 depicting a schematic flowchart of the method of the current application
  • Fig.3 depicting a graphical presentation of the net weight of stressed and control pepper plants during a period of one week, as monitored by the disclosed invention
  • Fig.4A depicting a graphical presentation of cumulative transpiration in a small-scale setting
  • Fig.4B depicting a graphical presentation of cumulative transpiration in a small-scale setting
  • Fig.5A depicting a graphical presentation of number of flowers in a small-scale setting
  • Fig.5B depicting a graphical presentation of number of flowers in a small-scale setting
  • Fig.6A depicting a graphical presentation of whole plants’ dry weight in a small-scale setting
  • Fig.6B depicting a graphical presentation of whole plants’ dry weight in a small-scale setting
  • Fig.7A depicting a graphical presentation of root dry weight in a small-scale setting
  • Fig.7B depicting a graphical presentation of root dry weight in a small-scale setting
  • Fig.8 depicting a graphical presentation of total yield in a large-scale setting
  • Fig.9 depicting a graphical presentation of number of fmits in a large-scale setting. Summary of the invention:
  • a computer implemented non-transitory algorithm configured to interpret the analyzed data sets, comprising steps of: i. storing the analyzed data sets in the system; ii. optionally retrieving previous stored analyzed data sets in the system; iii. assessing the analyzed data sets and the previously stored data sets characteristic of each parameter; and iv. issuing a predictive recommendation report comprising the treatment combinations for field crops of the same plant populations growing in large- scale settings, wherein the environmental conditions in the large-scale settings resemble the environmental conditions characteristic of the small-scale settings.
  • the plant populations are agricultural or horticultural plants, selected from a group consisting of monocotyledonous or dicotyledonous plants.
  • the plant population are selected from a group consisting of wild type species, cultivars, varieties, genotypes, genetically-modified plants, grafted plants, plants grown from primed, coated or embedded seeds and any combination thereof.
  • the abiotic stress is selected from a group consisting of heat, cold, drought, salinity, osmotic stress, exposure to pollutants, toxins or hazardous chemicals, and physical injuries or wounding and any combination thereof.
  • the predetermined physiological or phenological parameters are selected from a group consisting of plant height, plant weight, length of the roots, length of the stem, length of the leaves, length of the branches, length between the nodes, number of nodes, number of fruits, number of flowers or inflorescences, fruit weight, root weight, germination ability, biomass, transpiration rate, water use efficiency, branching, appearance of adventitious roots, color, leaf shape, woodiness, optical data, reflectance data, x-ray data, thermal emission, audio data, ultrasonic data, haptic data, chemical data, electric data, responsiveness or response to an applied stimulus and any combination thereof.
  • experiment conditions are selected from a group consisting of: exposing the plant population to abiotic stress, treating the plant populations during the experiment conducted in the small-scale settings, treating the plant populations before the experiment conducted in the small-scale settings, treating the seeds of the plant population before the experiment conducted in the small-scale settings, and any combination thereof.
  • the treating is a treatment selected from a group consisting of: exposing the plant populations or seeds thereof to chemicals, exposing the plant populations or seeds thereof to biological agents, treating the plant populations or seeds thereof with physical forces, and any combination thereof. It is another object of the present invention to disclose the method as described above, wherein, the plant populations grown in the small-scale settings are grown in lower numbers and for a shorter period of time compared to the plant populations grown in the large-scale settings.
  • the phenotyping system is configured to: (i) calculate linear correlations between stomatal conductance and transpiration to plant productivity; (ii) weight- the plants at various times a day; and (iii) control irrigation schedules and water quantities.
  • the computer implemented non transitory software medium is a cloud-based software configured to analyze data collected from the phenotyping system.
  • the predictive recommendation report comprises recommendations selected from a group consisting of: prediction of the performance of the plant population in large-scale settings under specific abiotic stress, prediction of the percentages of plants from the plant populations expected to survive the abiotic stress or perform better under the abiotic stress in large-scale settings, defining subpopulations of stress-tolerant plants to be further used for crossing, breeding and cultivation purposes, implementing management systems or protocols, taking cautionary actions and any combination thereof.
  • an artificial intelligence -based system for screening plant populations in small-scale settings and for predicting abiotic stress tolerance in the plant populations in large-scale settings, comprising: a. a phenotyping system; b. a computer implemented non transitory software medium; and c. a computer implemented non transitory algorithm; wherein the phenotyping system is configured to (i) monitor the plant populations under control and experiment conditions; and (ii) generate data sets, the computer implemented non transitory software medium is configured to (i) graphically and statistically analyze the data sets and (ii) result in analyzed data sets, and the computer implemented non transitory algorithm is configured to (i) interpret the analyzed data sets and (ii) generate a predictive recommendation report.
  • the plant populations are agricultural or horticultural plants, selected from a group consisting of monocotyledonous or dicotyledonous plants.
  • the abiotic stress is selected from a group consisting of heat, cold, drought, salinity, osmotic stress, exposure to pollutants, toxins or hazardous chemicals, and physical injuries or wounding and any combination thereof.
  • the experiment conditions are selected from a group consisting of: exposing the plant population to abiotic stress, treating the plant populations during the experiment conducted in the small-scale settings, treating the plant populations before the experiment conducted in the small-scale settings, treating the seeds of the plant population before the experiment conducted in the small- scale settings, and any combination thereof.
  • the phenotyping system is configured to (i) calculate linear correlations between stomatal conductance and transpiration to plant productivity, (ii) weigh the plants at various times a day and (iii) control irrigation schedules and water quantities.
  • the computer implemented non transitory software medium is a cloud-based software configured to analyze data collected from the phenotyping system.
  • the predictive recommendation report comprises recommendations selected from a group consisting of: prediction of the performance of the plant population in large-scale settings under specific abiotic stress, prediction of the percentages of plants from the plant populations expected to survive the abiotic stress or perform better under the abiotic stress in large-scale settings, defining subpopulations of stress-tolerant plants to be further used for crossing, breeding and cultivation purposes, implementing management systems or protocols, taking cautionary actions and any combination thereof.
  • Agricultural plants generally refers hereinafter to plants that are cultivated by humans for food, feed, and other industrials purposes, such as fiber and fuel.
  • Agricultural plants include both monocotyledonous species such as: maize ( Zea mays), common wheat ( Triticum aestivum), rice ( Oryza sativa ), and dicotyledonous species such as: pepper ( Capsicum annuum ), soybean ( Glycine max), canola and rapeseed cultivars ( Brassica napus), cotton (genus Gossypium), potato ( Solarium tuberosum), tomato ( Solarium ly coper sicum), pea ( Pisum sativum), chick pea ( Cicer arietinum) and many other varieties of vegetables.
  • plant population generally refers hereinafter to a group of plants whose ability to endure abiotic stress is evaluated and calculated in the present disclosure.
  • the population of plants may comprise wild-type plants, a specific genotype, germplasm, mutagenized plants, genetically modified plants, genome-edited plants, grafted plants, a specific variety, a specific cultivar, a specific species, plants exposed to a biological/chemical/physical treatment before or during the experiments disclosed in the present invention and more.
  • abiotic stress generally refers hereinafter to any condition to which plants are subjected, which is characterized by: (a) not being the optimal or natural conditions suitable for the plant’s growth and development; and (b) not caused by a living organism, such as nematodes, parasites, herbivores and the like.
  • abiotic stresses are for example heat, cold, drought, salinity, osmotic stress, exposure to pollutants, toxins or hazardous chemicals, and physical injuries.
  • the intracellular homeostasis becomes unbalanced, resulting in abnormal misfolding and aggregation of proteins, mitochondrial overload and excessive production of reactive oxygen species and free radicals.
  • salinity stress generally refers hereinafter to a specific type of abiotic stress, wherein a plant is exposed to abnormal or excessive amounts of salt in the soil or in the water.
  • Salinity stress affects plant growth and development via water stress, cytotoxicity due to an excessive uptake of ions, such as sodium and chloride, and nutritional imbalance. Additionally, salinity is typically accompanied by oxidative stress due to generation of reactive oxygen species and free radicals.
  • physiological/phenological/ performance parameters generally refers hereinafter to any observable, quantifiable or measurable external/internal characteristic of the plant, as recorded by the systems and computerized tools mentioned in the present application. Such characteristics can be related to, but are not limited to: height, weight, length of the roots, stem, leaves or branches, length between the nodes, number of nodes, number of fruits, number of flowers or inflorescences, germination, biomass, branching, appearance of adventitious roots, color, leaf shape, woodiness, roots’ weight, transpiration rate, water use efficiency and more.
  • small-scale settings generally refers hereinafter to a location or locations in which populations of plants are grown and tested for their performance and tolerance under various environmental conditions by the system and algorithm of the present invention. This may include for example laboratories or small greenhouses in academic or research centers’ facilities. The number of plants grown in those small-scale settings is smaller and the growth period is considered shorter in comparison to the number and period length of plants of the same population, grown in large scale settings, such as open fields. The plants grown in small scale population serve as the population on which experiments and testing are carried out, and their performance parameters are translated by the algorithm of the present invention to predictions and recommendations, concerning future growth of the same plant in large-scale settings.
  • the algorithm of the present application stores data from both small-scale and large scale setting experiments, thus, it is a self-learning algorithm, which is constantly updated. Conducting the experiments in small- scale settings saves time and money, as the growers do not need to wait for the plants’ complete growth cycle and results can be obtained on a relatively small number of plants, when the plants are still relatively young.
  • large-scale settings generally refers hereinafter to a location or locations in which populations of plants are grown following the indications and recommendations generated from data obtained from populations of the same plants grown in small-scale settings or from previous large-scale experiments, by the system and algorithm of the present invention.
  • Large-scale settings may include, in a non-limiting way fields, commercial greenhouses etc.
  • the plants grown in those large-scale settings may be grown to a full growth cycle, with the aim of resulting in produce and yield.
  • the populations of plants grown in those large-scale settings outnumber the populations of the plants grown in the small-scale settings, and they are grown for commercial purposes.
  • observation conditions generally refers hereinafter to the environmental conditions under which the physiological/phenological/ performance parameters of plant populations (which are not the control group(s)) are evaluated and calculated in small-scale settings. This may be for example, conditions of abiotic stress (such as high concentrations of salt in the soil/water, water deficiency, elevated temperatures etc.). In addition to mere environmental stressful conditions, the term “experiment conditions” may also mean other treatments carried out on the plant populations. For example, seeds can be treated before the small-scale setting experiment with plant hormones or horticultural chemicals, and subsequently be exposed to abiotic stress during the experiment. As a non-limiting example, the present disclosure describes populations of plants tested for their ability to endure salinity stress.
  • the experiment should be carried out in a research lab on a relatively small number of groups for a relatively short period of time (compared to a commercial crop cycle of the same plants in the field).
  • the plant populations may consist of a “control group” (group A) which is grown under optimal conditions (without excessive amounts of salt in the water), a group which is grown with excessive amounts of salt in the water (group B), another group whose seeds were pre-treated with a chemical, and grown with salty water during the experiment (group C) and an additional group, which is a salt-tolerant genotype (group D). Therefore, the term “experiment condition” refers to groups B-D.
  • the term “algorithm/ a computer implemented non-transitory algorithm/ computer processor” generally refers hereinafter to computer-implementable instructions and specifications designed to execute among other things calculations, data processing and task solving.
  • the disclosed algorithm of the present invention is specifically designed to integrate data obtained from small-scale setting experiments via a phenotypic screening system and a designated software which combined, capture images of the plants and calculates various performance parameters thereof.
  • the algorithm is fed with the above-mentioned data, retrieves previously stored data from past small-scale and large-scale experiments, correlates the environmental conditions of the experiment with the conditions in a large-scale setting, indicates plant populations/individual plants outperforming other plant populations, stores data from a current experiment, and generates a predictive recommendation report based on the data above.
  • the algorithm of the present invention is a self-learning algorithm, which is constantly updated by data obtained from both small-scale and large-scale experiments.
  • predictive recommendations/indications report generally refers hereinafter to data obtained from small-scale setting experiments or previous large-scale setting experiments concerning plant performance, physiological/phenological parameters or tolerance, which is analyzed by the algorithm of the present invention.
  • the data are translated to practical, helpful recommendations guiding and instructing growers for example, when to sow seeds in a field, which plant populations/genotypes are more likely to exhibit tolerance to abiotic stress in large- scale setting conditions, what is the percentage of plants within a specific population which would be more likely to survive and perform under stress, if preventive actions of any other crop management nature should be taken, etc.
  • Artificial intelligence (Al)-based system and a method for predicting abiotic stress tolerance and performance in tested plant populations.
  • the system provided herein comprises several key components to perform a supervised machine learning algorithm for identifying and/or measuring physiological and phenological parameters within a shorter time frame and on a smaller number of plants than evaluating an entire agricultural crop cycle after determining a significant stress condition by the disclosed AI-assisted system for plant stress mitigation.
  • the system comprises the following features:
  • the algorithm collects and evaluates the above data sets acquired by the automated plant screening complex for analyzing and predicting the performance of the current batch of plants in large-scale settings (such as fields) based on the previous data sets.
  • the data sets are constantly updated by data from ongoing screening in the field and in the lab, modifying the selection algorithm.
  • the present invention comprises methods for monitoring the physiological/phenological parameters of a selected group of plants grown under environmentally controlled conditions, to screen and select a plurality of plants at early stages of growth for their future phenotypes, analyze the obtained data and interpret it into practical recommendations and indications, for instance, which genotype is more likely to perform better under specific conditions once it is grown in large numbers for commercial purposes.
  • Pluralities of plants could be tested and analyzed by the system and algorithm of the present invention, including in a non-limiting fashion, nursery and wild type plants, as well as transgenic events or mutagenized plants which can be screened by the disclosed system.
  • the individual plants are exposed to a controlled environment, and automatically provided with controlled amounts of water, and/or nutrients based upon one or more assay definitions.
  • the disclosure comprises an automated method for phenotypic screening of a population of plants in small-scale settings using physiological/phenological parameters.
  • the method comprises subjecting the population of plants to an induced abiotic stress in a controlled environment and measuring the physiological/phenological parameters that vary in response to said stress.
  • An unstressed control and a stressed control are present in all the experiments, setting the upper and lower limits for the algorithms for selection.
  • the screening system and algorithm of the present invention monitors, collects and analyzes data regarding physiological/phenological parameters, performance and tolerance from populations of plants grown in small-scale settings.
  • plant species, number of tested plant groups, growth conditions and parameters may vary depending on the experiment and the growers’ needs and requirements
  • tomato plants Solanum lycopersicum
  • the different plant groups are grown and monitored under the surveillance of a phenotypic screening system, which takes measurements of various plants’ physiological/performance parameters.
  • the investigators wish to check if the various screened plant populations exhibit an increase in their biomass (weight) compared to the control group at day 28 from sowing. All the data are collected and analyzed by the algorithm of the present invention.
  • the output of said analysis comprises recommendations and indications for the next crop cycles under those field conditions, predicting how well the same tomato plants would perform in a field, which contains similar concentrations of salts compared with the lab experiment.
  • the investigators compare in laboratory settings different genotypes/germplasm of tomato plants grown under the same salt concentration.
  • the data obtained from such small-scale experiment are analyzed and result in a report which specifies which genotype is more tolerant to salinity stress and would be more likely to perform better in a field characterized by having similar salinity conditions.
  • the disclosed invention simulates the field conditions or large-scale setting on a smaller, lab scale setting and screens the plant populations best suited for that large-scale setting in shorter time.
  • the growers might grow the plants in a large-scale setting (such as a field) having similar environmental conditions as the laboratory experiment and evaluate if the predictive recommendation and indication report is accurate and helpful.
  • the growers report back and give their feedback, so that actual results from large-scale settings are fed into the algorithm of the present invention, ensuring it is constantly updated and capable of retrieving said data in future experiments.
  • the disclosed screening system and algorithm are configured to capture, takes measurements or calculate various physiological/phenological/performance parameters and to present them as an output graphically, visually, statistically, schematically, illustratively or by any other mean acceptable or known in the art.
  • the disclosed screening system and algorithm are configured to compare different plant populations, and additionally carry out comparisons, measurements and calculations within the same population of plants. For instance, when growing and monitoring the plants in a small-scale setting, the screening system and the algorithm may indicate and calculate how many plants from a specific group perform better or are more tolerant to the stress.
  • the disclosed screening system and algorithm are configured to integrate multiple variants in parallel.
  • the phenotypic screening system is configured to monitor a plurality of physiological/phenological/performance parameters simultaneously on numerous plant populations and extract data concerning all the desired parameters. For instance, the system can monitor at the same time a plurality of plant populations and extract data concerning the plant height, transpiration rate, number of fruits/flowers per plant, plant weight, number of leaves and calculate all those data for each individual plant.
  • the disclosed screening system and algorithm are configured to evaluate different performance parameters in small-scale settings, and determine if those parameters are appropriate and indicative for the designed experiment and the actual, subsequent commercial growth of the plants.
  • the scientific literature is replete with information and data regarding which plant performance parameters should be taken into consideration under abiotic stress. Nevertheless, not all these data or the data obtained from small- scale experiments properly correlate with plant performance in the field.
  • the algorithm of the present invention is constantly updated by new results from small-scale settings, real time results and feedback from the large-scale settings and scientific publications, it can filter out parameters which are considered by the scientific literature as reliable or accurate, or parameters considered promising and predictive in past small-scale experiments, in case those parameters did not materialize nor led to satisfactory results in the field.
  • the algorithm would flag it out in the predictive recommendation report, and highlight other predictive performance parameters, which bear greater significance.
  • the algorithm of the present invention analyzes and integrates a plurality of parameters and data sets obtained from individual plants, and is also a self-learning algorithm, configured to retrieve and assess previous large-scale performance data, the resultant predictive recommendation report is highly accurate.
  • the disclosed algorithm collects and stores data from past small-scale and large-scale setting experiments, and retrieves them for present analyses.
  • the data collected for example, from past large-scale experiments may include in a non-limiting way: yield parameters, produce export quality, characteristics of the salt concentrations in the soil or water and more.
  • growers and seed companies are provided with helpful, useful data in the form of predictive recommendations to determine the ability of agricultural crops to mitigate abiotic stress conditions in the field.
  • the analyzed data obtained from the screening system and the algorithm of the present invention may assist growers to better understand when it is productive to initiate a crop management action, as opposed to when it may not be productive to take actions.
  • the system, method and algorithm disclosed herein can save time and costs. As the data is obtained from a relatively small number of plants (compared to the commercial quantities required in a field), and sometime there is no need to wait for the end of the growth cycle in order to obtain the desired data, the present invention can save time and money for growers and seed companies, otherwise spent on large-scale experiments, potentially involving substantial yield loss.
  • the present invention provides an automated AI-based screening system for abiotic stress mitigation in plants which is also predictive of abiotic stress tolerance in large-scale settings.
  • the system used in the present set of experiments is designed as one-to-one (1:1) plant- [sensors+controller] unit, i.e., each individual plant is monitored by one unit of sensor, controller and irrigation valves that enable: (i) monitoring water-relation kinetics of each plant and environmental responses throughout the plant’s life cycle with high spatiotemporal resolution, (ii) creating a truly randomized experimental design due to multiple independent treatment scenarios for every plant, and (iii) reducing artificial ambient perturbations due to the immobility of the plants or other objects.
  • Plantarray 3.0 platform of Plant-Ditech (www.plant- ditech.com/products/plantarray) is an exemplary system which may be used for monitoring plants’ performance during the entire experimental period (in small-scale settings) by controlling irrigation schedule and water quantities.
  • This platform enables performing high-throughput physiological functional phenotyping by continuously, simultaneously and accurately measuring the momentary-varying water flux in the soil-plant atmosphere for each plant in the array.
  • the Plantarray phenotypic system is based on the linear correlation between stomatal conductance and transpiration to plant productivity (i.e. CO2 assimilation), which indicates the plant performance in high correlation to yield results.
  • the system includes up to 72 units of highly sensitive, temperature-compensated load cells that are used as weighing lysimeters. Each unit is connected to a personalized controller, which measures the pot weight 24/7, collects the data and controls the irrigation of each plant separately. An independent controller for each pot enables tight feedback irrigation system, based on the plant’s transpiration rate. Each controller unit is connected to its neighboring unit for serial data collection and all data are loaded to a server. A pot with three plants is placed on each load cell.
  • the data can be analyzed by a designated software, such as SPACanalytics by Plant-Ditech (www.plant-ditech.com/products/spac-analytics), an online cloud- based software that enables viewing and graphically and statistically analyzing the real-time data collected from the Plantarray system.
  • the estimated plant’s weight at the beginning of the experiment is calculated as the difference between the total system weight and the sum of the tare weight of pot + drainage container, weight of soil at pot capacity, and weight of water in the drainage container at the end of the free drainage.
  • the plant’s weight at the end of a growth period is calculated as the sum of the initial plant’s weight and the multiplication of the cumulative transpiration during the period by the water use efficiency (WUE).
  • WUE water use efficiency
  • plant data such as optical data, reflectance data, x-ray data, thermal emission, audio data (e.g., ultrasonic data, etc.), haptic data, chemical data (e.g., chemical composition of the plant), electric data (e.g., resistivity, voltage open circuit, inductance, electrical noise, conductance, etc.), responsiveness or response to an applied stimulus (e.g., incident light, acoustic noise, haptic stimulus, electric stimulus), thermal data, could form a data set in the system.
  • the collected signals can be within the range of human detection, but can alternatively or additionally be determined based on signals outside the range of human detection.
  • measurements can be taken using or recording an audio frequency outside the aural frequency range or a light frequency outside the visual spectrum.
  • the plant data can additionally or alternatively be chemical characteristics (e.g., chemical composition, concentration of a given chemical, etc.), visual characteristics, electrical characteristics, or any other suitable characteristic of the plant.
  • the plant data are preferably collected from the entirety of the plant body, but can alternatively be collected from a portion (e.g., less than the entirety) thereof, such as a leaf, a flower or a fruit.
  • the plant data are preferably non-destructively collected from the plants, but can alternatively be destructively collected from the plant.
  • the plant data are preferably collected over multiple sessions, spaced over a period of time, such that the plant characteristics are tracked across time. However, the plant data can be collected at any other suitable frequency, for instance twice a day or 3 times a week.
  • the tested plant populations or any part thereof can be treated during or before the small-scale settings experiments with various chemicals or biochemical compounds, such as plant hormones, lipids, peptides, proteins PEG, reactive oxygen species etc., biological agents, such as bacteria, fungi, viruses etc., or physical treatments such as irradiation, contact forces, electrical forces and more.
  • various chemicals or biochemical compounds such as plant hormones, lipids, peptides, proteins PEG, reactive oxygen species etc., biological agents, such as bacteria, fungi, viruses etc., or physical treatments such as irradiation, contact forces, electrical forces and more.
  • Plant data can be utilized to automatically select plants that express desired phenotypes from the plurality of plants within the plant field.
  • the identified plants are preferably used for successive breeding, wherein the genetic material of the selected plants is preferably used in the creation of the next generation of plants (e.g., by sexual reproduction, genetic vector methods, etc.).
  • the identified plants can be asexually reproduced, such as by taking a cutting and grafting the cutting onto a host plant or by using stem or tuber cuttings.
  • the phenotypic data can additionally or alternatively be used to test individual plant reactions to given treatments or management systems, wherein specific plant phenotypes that are susceptible (or conversely, resistant) to the given treatment can be identified and selected from a field of plants. Individual plant responses to given treatments can additionally be used to make calculated, judicious future planting or cropping decisions.
  • the selected plants manifesting improved traits in small-scale setting experiments, such as stress tolerance, as detected by the disclosed system and algorithm, can be subsequently recommended to growers to be planted in environments having predicted parameters similar to the microclimate in which the tested plants were grown and evaluated.
  • the selected plants can alternatively be subsequently recommended to growers to be planted with management systems similar to the plant management system and/or plans used for the tested plant population in said small-scale settings.
  • the plants to be screened in the system are preferably agricultural crops, but can alternatively be any other suitable plants.
  • the tested populations could comprise proprietary/ non-proprietary gene pool materials, recombinant DNA products, plants grown from primed, coated or embedded seeds, products of mutagenesis or any other genetic or molecular manipulations.
  • the system 100 of the present disclosure is depicted in Fig. 1.
  • a grower/seed company is interested in evaluating and predicting the ability of a specific variety/cultivar/genotype/germplasm of plants to grow and perform in a place, where the soil is known to contain excessive amounts of salt. Seeds of these plants are sown in a research laboratory, in a controlled environment mimicking the conditions of said field.
  • the plants grown in the lab comprise two groups: a control group (plants growing under natural and optimal conditions without stress) and a stress group (plants exposed to high concentrations of salt).
  • the plant populations 101 are continuously monitored by a monitoring platform 102 (for instance, the phenotyping Plantarray 3.0 platform) for a defined period of time. Each plant is weighed 100-150 times within 24 hours for 10 days, and the amounts of water used for irrigation and irrigation timing are strictly recorded and monitored.
  • the disclosed system 100 collects data regarding phenotypical/phenological parameters from the plant populations (both control and stress groups) and generates corresponding databases 103. After collecting the phenotypical/phenological data sets, the system 100 utilizes a graphical user interface 104 (for example, the cloud-based SPACanalytics software) to generate graphs, calculations and statistical analyses pertinent to the performance of the plant population 101 under salinity stress.
  • a graphical user interface 104 for example, the cloud-based SPACanalytics software
  • the system 100 employs a computer processor 105 (the algorithm of the disclosed application) to: (a) store all the data sets collected from the current experiment in the system’s databases 103; and (b) integrate the collected data sets and convert the graphical and statistical information generated by the graphical user interface 104 into a predictive recommendation report for growers 107, provided the system 100 detects individual plants which perform better under the defined salinity stress. If such tolerant individual plants are detected 106 within the entire plant populations 101, the disclosed system 100 issues a predictive recommendation report 107.
  • a computer processor 105 the algorithm of the disclosed application
  • the computer processor 105 is configured to also access and utilize other optional databases 108, such as data collected and stored from previous small-scale and large-scale settings experiments performed by the system 100, or known data obtained from publicly available resources, such as publications and scientific literature.
  • the predictive recommendation report 107 can comprise for instance, predictions of the percentage of plants from the specific tested variety/cultivar/genotype which would grow and perform well under similar field conditions, recommendations to cross individual salt-tolerant plants with sensitive plants or designing management plans which would assist in the technical maintenance of the plants, should they encounter similar stress conditions in the field.
  • the method 200 of the present disclosure is depicted in Fig. 2.
  • the disclosure commences when the need to predict plants’ tolerance to abiotic stress in the field arises.
  • the goal of the disclosed method is to screen relatively small populations of plants in a relatively short period of time and detect individual plants which are more resilient to stress or exhibit enhanced physiological/phenological parameters.
  • plants of a specific and defined plant populations are grown in a small-scale setting, such as a research laboratory under controlled conditions 201.
  • the plants populations are subjected to control (optimal, stress-free conditions) and to stressful conditions, the likes of which, the same plant populations will experience when commercially grown in large-scale settings, such as fields.
  • the seeds of these plants can be exposed to different treatments prior to sowing, to assess the effect of treatment on stress tolerance.
  • a special phenotyping platform continuously monitors the plants and takes measurements thereof (such as weight, height and transpiration rate) 202, and creates data sets for each plant population 203 (control and stress). Subsequently, a designated software is utilized to graphically and statistically analyze the data obtained from the experiments 204, and then, the special algorithm of the present invention is employed for the interpretation of the data sets 205, by integrating them, analyzing them and selecting the plant populations/individual plants which are likely to better perform and endure abiotic stress in a large-scale setting.
  • the algorithm of the disclosed method is also designed to store the data sets from the current small-scale experiment in the system for future purposes (for example, if a grower will be interested to grow the same type of plants under the tested microclimate, he/she could rely on the current results, predicting the plants’ ability to survive such conditions) 206. Additionally, the algorithm of the present invention is configured to store past data from small-scale setting experiments, large-scale setting experiments and scientific publications, and retrieving them if necessary. More particularly, the disclosed method comprises a step of issuing a predictive recommendation report 207 for the growers, containing all the necessary analyzed data collected during the experiments, and especially single out individual plants/plant populations, whose performance under stress is improved compared to other plants. By using the indications and recommendation of said report, growers would be able to manage their crop-growing more wisely and to select specific plants, which would perform better under abiotic stress in a field, thus resulting in increased yield.
  • VWC volumetric water content
  • the Physiological Phenotyping Platform was conducted in Aug-Sept 2019 in a commercial-like greenhouse (defined in the present disclosure as a “small-scale setting”) located at the Faculty of Agriculture, Food and Environment in Rehovot, Israel.
  • the greenhouse temperature was controlled using fans that blew air through a moist mattress, keeping it below 30°C.
  • the temperature and relative humidity (RH) were 30-37°C and 2-3.5%, respectively.
  • the plants were grown under natural light (midday maximum of 1100 pmol s-1 m-2), representative values for natural conditions during the summer in the central part of Israel, including Rehovot.
  • the temperature, RH, photosynthetically active radiation, barometric pressure and vapor pressure deficit in the greenhouse were continuously monitored by the meteorological station.
  • the pepper plants were monitored by the Plantarray 3.0 platform, which continuously took measurements of each individual plant.
  • One of the physiological parameters that was being evaluated was the effect of salinity stress on the plant net weight, as depicted in Fig. 3.
  • the plants were divided to the five following populations to be screened:
  • A) Unstressed control - plants growing under optimal/natural conditions, without exposure to any stress; and the following groups which comprise the “experiment conditions”: B) Stressed control - plants exposed to salinity stress (EC 4.7 - 3000 mg/L);
  • the algorithm will issue a predictive recommendation report to growers, indicating which seeds, when treated with a specific treatment, can grow better in salt-enriched soil, or which individual resistant plants from a certain population can be crossed with less tolerant varieties/cultivars in order to generate more salt resistant pepper plants.
  • the algorithm of the present invention is configured to store the data collected from this present small-scale experiment and to retrieve it for future analyses, if necessary.
  • the small-scale experiment comprised 5 populations of chemically treated plants (“T3 ec 4”; “T5 ec 4”; “T6 ec 4”; “T7 ec 4”; and “T8 ec 4”) and two control populations - “Ctrl ec 1” (unstressed and untreated) and “Ctrl ec 4” (salt-stressed and untreated).
  • the seeds were sown in a tray with 10-mL cones filled with commercial growing medium (Matza Gan, Shaham, Givat-Ada, Israel), composed of (w/w) 55% peat, 20% tuff and 25% puffed coconut coir fiber.
  • the trays were well irrigated and kept in the same greenhouse (on side-tables) where the experiment was performed.
  • the seedlings were 4 weeks old, the growing medium was carefully washed off (to avoid root damage) the seedling roots and the seedlings were immediately transferred to 4-L pots filled with 20/30 sand (Negev Industrial Minerals Ltd., Israel).
  • the functional phenotyping system Plantarray 3.0 platform (Plant-Ditech) was used to monitor the plants' performance during the entire experimental period by controlling the schedule and quantity of irrigation.
  • This platform which enables performing high-throughput physiological functional phenotyping, includes 72 units of highly sensitive, temperature-compensated load cells that are used as weighing lysimeters. Each unit is connected to its personalized controller, which collects the data and controls the irrigation of each plant separately. An independent controller for each pot enables tight feedback irrigation, based on the plant’s transpiration rate.
  • Each controller unit is connected to its neighbor for serial data collection and loading to a server. A pot with a single plant is placed on each load cell.
  • the data were analyzed by SPAC-analytics (Plant-Ditech), a designated online web-based software that enables viewing and analyzing the real-time data.
  • SPAC-analytics Plant-Ditech
  • the data obtained by the Plantarray 3.0 system indicated that one group of plants (referred to herein as “T5 ec 4”) which was subjected to salinity stress exhibited better physiological parameters than its respective control group (“ctrl ec 4”).
  • the plant group named herein “T7 ec 4” showed negative growth and development trends compared with the control group.
  • the following tables and figures show several parameters tested and measured by the disclosed system.
  • Table 1 shows cumulative transpiration of the different plant groups, with “T5 ec 4” exhibiting higher values than the control group, and “T7 ec 4” exhibiting lower values compared to control.
  • Fig. 4A graphically depicting differences in cumulative transpiration in control groups “ctrl ec 1” and “ctrl ec 4” compared to “T5 ec 4”
  • Fig. 4B graphically depicting differences in cumulative transpiration in control groups “ctrl ec 1” and Ctrl ec 4 compared to “T7 ec 4”.
  • An additional parameter evaluated by the system disclosed in the present invention is the number of flowers, a parameter known to be negatively affected by salinity stress. As demonstrated in Table 2 and graphically depicted in Fig. 5A and 5B, the pepper plants in group “T5 ec 4” had more flowers than its respective control group (37%), whereas group “T7 ec 4”, exhibited a reduction in the number of flowers compared to control (by 13%).
  • Table 2 Data on number of flowers in small-scale setting experiments The dry weight of the whole plant was also monitored and calculated by the system disclosed herein, as part of the small-scale experiment. As demonstrated in Table 3 and graphically depicted in Fig. 6A and 6B, the pepper plants in group “T5 ec 4” exhibited higher dry weight values than control group (19%), whereas group “T7 ec 4”, exhibited a lower dry weight compared to control (by 20%).
  • Table 4 demonstrates the differences between control group (“Ctrl ec 4”) to tested groups “T5 ec 4” and “T7 ec 4”.
  • the roots of group “T5 ec 4” exhibit an increase of 13% in weight compared to control group, whereas the roots of group “T7 ec 4” weigh less than the roots of control group (by 22%).
  • Fig. 7A graphically depicting differences in roots’ dry weight between control group and “T5 ec 4”.
  • Fig. 7B graphically depicts differences in roots’ dry weight between control group and “T7 ec 4”.
  • Table 4 - Data on roots’ dry weight in small-scale setting experiments After obtaining all the data above, the algorithm of the present invention is deployed.
  • the algorithm is designed to integrate all the data concerning the different plant populations’ performance under said salinity stress, analyze the data and generates a predictive recommendation report for potential growers.
  • the algorithm also retrieves data from previous laboratory and field experiments, as well as from scientific publications, and integrate said data with the data of the current experiment.
  • the report indicates that once grown in large-scale settings (such as fields for commercial purposes), with similar environmental conditions to those of the small- scale setting experiment, plants of the “T5 ec 4” group would be likely to perform better and to result in increased yield despite the abiotic stress.
  • the report would indicate that plants of the “T7 ec 4” group would result in decreased performance, and therefore, it would be disadvantageous and ill-advised for growers to plant them in this specific environment.
  • the experiment’s matrix comprised: the same 5 populations ( “T-3”; “T-5; “T-6”; “T-7”; and “T- 8”) monitored and screened in the small-scale experiment and a control group (referred to herein as “ctrl”), which is also stressed by the environmental conditions of the field, but untreated.
  • the populations were planted in randomized blocks - 4 repetitions, each repetition included 35 plants, where the middle 20 plant are the actual repetition.

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Genetics & Genomics (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Medical Informatics (AREA)
  • Mathematical Physics (AREA)
  • Evolutionary Computation (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Botany (AREA)
  • Developmental Biology & Embryology (AREA)
  • Environmental Sciences (AREA)
  • Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)

Abstract

Provided herein an artificial intelligence-based method for screening a plurality of plant populations and predicting plants' performance and yield under abiotic stress under field conditions, comprising the steps of: (1) sowing seeds of said plant populations in controlled small-scale settings under control and experiment conditions; (2) monitoring n combinations of predetermined physiological or phenological parameters of each plant of said plurality of plant populations under said control and experiment conditions over a predetermined period of time; (3) obtaining n data sets comprising values for said plurality of plant populations based on said predetermined physiological or phenological parameters; (4) analyzing said data sets graphically and statistically using a computer implemented non-transitory software medium having instructions to obtain analyzed data sets; and (5) employing a computer implemented non-transitory algorithm configured to interpret said analyzed data sets, comprising steps of: (i) storing said analyzed data sets in said system; (ii) optionally retrieving previous stored analyzed data sets in said system; (iii) assessing said analyzed data sets and said previously stored data sets characteristic of said each parameter; and (iv) issuing a predictive recommendation report comprising the treatment combinations for field crops of said same plant populations growing in large- scale settings, wherein the environmental conditions in said large-scale settings resemble the environmental conditions characteristic of said small-scale settings.

Description

SCREENING SYSTEM FOR ABIOTIC STRESS MITIGATION IN PLANTS
Field of the Invention
The present disclosure relates to a system and methods for automated Al-based phenotypic screening of plants, particularly under abiotic stress, and predicting plant performance in the field.
Background of the invention
The ever-growing world population, the climate crisis with its potentially devastating consequences to crops and the increasingly scarce supply of arable land available for agriculture are driving research to achieve greater efficiency in the field of agriculture. Marginal and suboptimal growing conditions, or stressful conditions, such as heat, drought, salinity, exposure to pollutants and chemicals etc., are considered a major challenge for growers and a great deal of efforts, research and resources have been invested in recent decades in order to optimize crop cultivation. Conventional means for horticultural and crop improvement comprise selective breeding techniques for the identification of plants that have desirable traits. However, such breeding techniques have several disadvantages, as they are typically labor intensive and require long periods of observations and evaluations, hence rendering the selection of exceptional individual plants complicated and time-consuming. Additionally, the gap between results obtained from laboratory and small-scale experiments and the actual plants’ performance and growth in the fields with their ever-changing environmental parameters and factors is still relatively great, and more sophisticated and accurate tools and means must be developed.
There have been a number of systems for screening of crop plants for desired traits, based on physiological parameters such as canopy temperature, biomass or seedling vigor.
US patent 9025831 discloses an apparatus for screening potted plants and an automated method for high-throughput phenotypic screening of a plurality of plants. The method comprises growing a population of plants, which may comprise one or more transgenic events, in a controlled environment for a defined growing period and are subjected to at least one induced stress. The growth profile comprises at least three measurable characteristics selected from the group consisting of canopy area, chlorophyll content, anthocyanin content, biomass, plant height, root mass, water content, yield, the amount of water applied during the growing period, and water use efficiency. After the growth profile is determined, individual plants can be selected from the population of plants based upon the determined growth profile.
Patent application WO2014/124128 discloses systems and methods for plant stress mitigation. The computer-implemented system, which in first aspect, provides a self-standardizing algorithm that can be universally applied to detect stress-related changes in plants stress before it is visible to the naked eye and objectively quantify that stress over time and space.
US patent 10492374 discloses a method for acquiring data associated with a plant growing in a field using a sensor, analyzing the data obtained from the sensor to extract, while plants still grow in the field, one or more phenotypes, determining one or more plant insights based on phenotypic traits, including information about the plant’s health, yield, growth, harvest, management, performance, or state; and generating a plant insights report. Furthermore, the analysis of the data obtained from the sensor includes: predicting the phenotypic traits based on the sensor data and a computerized model; displaying indications of the phenotypic traits predicted; and obtaining a confirmation, modification, or addition indication from the user for at least one of the indications of the phenotypic traits predicted based on direct observation of the plant in the field by the user.
US Patent application 2011/0125477A1 discloses methods and related devices for predicting the presence or level of one or more characteristics of a plant or plant population based on spectral, multi-spectral, or hyper- spectral data obtained by, e.g., remote sensing. The predictions and estimates furnished by the inventive methods and devices are useful in crop management, crop strategy, and optimization of agricultural production. The method of estimating a plant characteristic comprises: 1) building a predictive model using inverse modeling using: i. a first set of spectroscopic data from a first plant population, and ii. corresponding measured characteristic data sets from the first plant population; and, 2) applying the model to a second set of spectroscopic data from a second plant, a second plant population, or both, so as to estimate the characteristic in the second plant.
In light of the prior art documents, there is still an unmet need for an artificial intelligence-based system capable of (a) screening populations of plants in laboratory/small-scale settings for phenotypic differences and traits, under various environmental conditions or treatments, (b) performing analyses based on the data gathered from the laboratory experiments, and (c) translating said data into practical actions and recommendations for growers and seed companies by a unique algorithm, which can predict how individual plants/genotypes will perform in the field overtime under specific conditions, including abiotic stresses. The present invention discloses such a system and algorithm configured to monitor plant performance under abiotic stress in early stages of crops in a relatively short period of time, and at relatively small number of plants and generate corresponding predictive recommendations and indications for growers.
Brief description of the figures
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention.
Fig.l depicting a schematic flowchart of the system of the present application;
Fig.2 depicting a schematic flowchart of the method of the current application;
Fig.3 depicting a graphical presentation of the net weight of stressed and control pepper plants during a period of one week, as monitored by the disclosed invention; Fig.4A depicting a graphical presentation of cumulative transpiration in a small-scale setting;
Fig.4B depicting a graphical presentation of cumulative transpiration in a small-scale setting;
Fig.5A depicting a graphical presentation of number of flowers in a small-scale setting;
Fig.5B depicting a graphical presentation of number of flowers in a small-scale setting;
Fig.6A depicting a graphical presentation of whole plants’ dry weight in a small-scale setting; Fig.6B depicting a graphical presentation of whole plants’ dry weight in a small-scale setting;
Fig.7A depicting a graphical presentation of root dry weight in a small-scale setting;
Fig.7B depicting a graphical presentation of root dry weight in a small-scale setting;
Fig.8 depicting a graphical presentation of total yield in a large-scale setting; and
Fig.9 depicting a graphical presentation of number of fmits in a large-scale setting. Summary of the invention:
It is an object of the present invention to disclose an artificial intelligence-based method for screening a plurality of plant populations and predicting plants’ performance and yield under abiotic stress under field conditions, comprising the steps of: a. sowing seeds of the plant populations in controlled small-scale settings under control and experiment conditions; b. monitoring n combinations of predetermined physiological or phenological parameters of each plant of the plurality of plant populations under the control and experiment conditions over a predetermined period of time; c. obtaining n data sets comprising values for the plurality of plant populations based on the predetermined physiological or phenological parameters; d. analyzing the data sets graphically and statistically using a computer implemented non- transitory software medium having instructions to obtain analyzed data sets; and e. employing a computer implemented non-transitory algorithm configured to interpret the analyzed data sets, comprising steps of: i. storing the analyzed data sets in the system; ii. optionally retrieving previous stored analyzed data sets in the system; iii. assessing the analyzed data sets and the previously stored data sets characteristic of each parameter; and iv. issuing a predictive recommendation report comprising the treatment combinations for field crops of the same plant populations growing in large- scale settings, wherein the environmental conditions in the large-scale settings resemble the environmental conditions characteristic of the small-scale settings.
It is another object of the present invention to disclose the method as described above, wherein, the plant populations are agricultural or horticultural plants, selected from a group consisting of monocotyledonous or dicotyledonous plants.
It is another object of the present invention to disclose the method as described above, wherein, the plant population are selected from a group consisting of wild type species, cultivars, varieties, genotypes, genetically-modified plants, grafted plants, plants grown from primed, coated or embedded seeds and any combination thereof. It is another object of the present invention to disclose the method as described above, wherein, the abiotic stress is selected from a group consisting of heat, cold, drought, salinity, osmotic stress, exposure to pollutants, toxins or hazardous chemicals, and physical injuries or wounding and any combination thereof. It is another object of the present invention to disclose the method as described above, wherein, the predetermined physiological or phenological parameters are selected from a group consisting of plant height, plant weight, length of the roots, length of the stem, length of the leaves, length of the branches, length between the nodes, number of nodes, number of fruits, number of flowers or inflorescences, fruit weight, root weight, germination ability, biomass, transpiration rate, water use efficiency, branching, appearance of adventitious roots, color, leaf shape, woodiness, optical data, reflectance data, x-ray data, thermal emission, audio data, ultrasonic data, haptic data, chemical data, electric data, responsiveness or response to an applied stimulus and any combination thereof.
It is another object of the present invention to disclose the method as described above, wherein, the electric data are selected from a group consisting of resistivity, voltage open circuit, inductance, electrical noise, conductance and any combination thereof.
It is another object of the present invention to disclose the method as described above, wherein, the experiment conditions are selected from a group consisting of: exposing the plant population to abiotic stress, treating the plant populations during the experiment conducted in the small-scale settings, treating the plant populations before the experiment conducted in the small-scale settings, treating the seeds of the plant population before the experiment conducted in the small-scale settings, and any combination thereof.
It is another object of the present invention to disclose the method as described above, wherein, the treating is a treatment selected from a group consisting of: exposing the plant populations or seeds thereof to chemicals, exposing the plant populations or seeds thereof to biological agents, treating the plant populations or seeds thereof with physical forces, and any combination thereof. It is another object of the present invention to disclose the method as described above, wherein, the plant populations grown in the small-scale settings are grown in lower numbers and for a shorter period of time compared to the plant populations grown in the large-scale settings.
It is another object of the present invention to disclose the method as described above, wherein, the phenotyping system is configured to: (i) calculate linear correlations between stomatal conductance and transpiration to plant productivity; (ii) weight- the plants at various times a day; and (iii) control irrigation schedules and water quantities.
It is another object of the present invention to disclose the method as described above, wherein, the computer implemented non transitory software medium is a cloud-based software configured to analyze data collected from the phenotyping system.
It is another object of the present invention to disclose the method as described above, wherein, the predictive recommendation report comprises recommendations selected from a group consisting of: prediction of the performance of the plant population in large-scale settings under specific abiotic stress, prediction of the percentages of plants from the plant populations expected to survive the abiotic stress or perform better under the abiotic stress in large-scale settings, defining subpopulations of stress-tolerant plants to be further used for crossing, breeding and cultivation purposes, implementing management systems or protocols, taking cautionary actions and any combination thereof.
It is also the object of the present invention to disclose an artificial intelligence -based system for screening plant populations in small-scale settings and for predicting abiotic stress tolerance in the plant populations in large-scale settings, comprising: a. a phenotyping system; b. a computer implemented non transitory software medium; and c. a computer implemented non transitory algorithm; wherein the phenotyping system is configured to (i) monitor the plant populations under control and experiment conditions; and (ii) generate data sets, the computer implemented non transitory software medium is configured to (i) graphically and statistically analyze the data sets and (ii) result in analyzed data sets, and the computer implemented non transitory algorithm is configured to (i) interpret the analyzed data sets and (ii) generate a predictive recommendation report.
It is another object of the present invention to disclose the system as described above, wherein, the plant populations are agricultural or horticultural plants, selected from a group consisting of monocotyledonous or dicotyledonous plants.
It is another object of the present invention to disclose the system as described above, wherein, the plant population are selected from a group consisting of wild type species, cultivars, varieties, genotypes, genetically-modified plants, grafted plants, plants grown from primed, coated or embedded seeds and any combination thereof.
It is another object of the present invention to disclose the system as described above, wherein, the abiotic stress is selected from a group consisting of heat, cold, drought, salinity, osmotic stress, exposure to pollutants, toxins or hazardous chemicals, and physical injuries or wounding and any combination thereof. It is another object of the present invention to disclose the system as described above, wherein, the experiment conditions are selected from a group consisting of: exposing the plant population to abiotic stress, treating the plant populations during the experiment conducted in the small-scale settings, treating the plant populations before the experiment conducted in the small-scale settings, treating the seeds of the plant population before the experiment conducted in the small- scale settings, and any combination thereof.
It is another object of the present invention to disclose the system as described above, wherein, the phenotyping system is configured to (i) calculate linear correlations between stomatal conductance and transpiration to plant productivity, (ii) weigh the plants at various times a day and (iii) control irrigation schedules and water quantities. It is another object of the present invention to disclose the system as described above, wherein, the computer implemented non transitory software medium is a cloud-based software configured to analyze data collected from the phenotyping system.
It is another object of the present invention to disclose the system as described above, wherein, the predictive recommendation report comprises recommendations selected from a group consisting of: prediction of the performance of the plant population in large-scale settings under specific abiotic stress, prediction of the percentages of plants from the plant populations expected to survive the abiotic stress or perform better under the abiotic stress in large-scale settings, defining subpopulations of stress-tolerant plants to be further used for crossing, breeding and cultivation purposes, implementing management systems or protocols, taking cautionary actions and any combination thereof.
Detailed description of the preferred embodiments
The term “about” generally refers hereinafter to a measurable value such as an amount, a temporal duration, and the like, meant to encompass variations of +/- 20%, more preferably +/- 5%, even more preferably +/- 1%, and still more preferably +/- 0.1% from the specified value, as such variations are appropriate to perform the disclosed methods.
The term "plurality" as used herein, means two or more.
The term "agricultural plants/crops" generally refers hereinafter to plants that are cultivated by humans for food, feed, and other industrials purposes, such as fiber and fuel. Agricultural plants include both monocotyledonous species such as: maize ( Zea mays), common wheat ( Triticum aestivum), rice ( Oryza sativa ), and dicotyledonous species such as: pepper ( Capsicum annuum ), soybean ( Glycine max), canola and rapeseed cultivars ( Brassica napus), cotton (genus Gossypium), potato ( Solarium tuberosum), tomato ( Solarium ly coper sicum), pea ( Pisum sativum), chick pea ( Cicer arietinum) and many other varieties of vegetables.
The term “plant population” generally refers hereinafter to a group of plants whose ability to endure abiotic stress is evaluated and calculated in the present disclosure. The population of plants may comprise wild-type plants, a specific genotype, germplasm, mutagenized plants, genetically modified plants, genome-edited plants, grafted plants, a specific variety, a specific cultivar, a specific species, plants exposed to a biological/chemical/physical treatment before or during the experiments disclosed in the present invention and more.
The term “abiotic stress” generally refers hereinafter to any condition to which plants are subjected, which is characterized by: (a) not being the optimal or natural conditions suitable for the plant’s growth and development; and (b) not caused by a living organism, such as nematodes, parasites, herbivores and the like. Among abiotic stresses are for example heat, cold, drought, salinity, osmotic stress, exposure to pollutants, toxins or hazardous chemicals, and physical injuries. During abiotic stress the intracellular homeostasis becomes unbalanced, resulting in abnormal misfolding and aggregation of proteins, mitochondrial overload and excessive production of reactive oxygen species and free radicals. The above molecules can become deleterious, as they tend to be extremely reactive, oxidize and break down proteins, membranes, fatty acids and nucleic acids. Naturally, not all plants experience abiotic stress in a similar way and the severity of the stress is determined, among other things, by its length and by the plant’s species, size and stage (reproductive or vegetative).
The term “salinity stress” generally refers hereinafter to a specific type of abiotic stress, wherein a plant is exposed to abnormal or excessive amounts of salt in the soil or in the water. Salinity stress affects plant growth and development via water stress, cytotoxicity due to an excessive uptake of ions, such as sodium and chloride, and nutritional imbalance. Additionally, salinity is typically accompanied by oxidative stress due to generation of reactive oxygen species and free radicals.
The term “physiological/phenological/ performance parameters” generally refers hereinafter to any observable, quantifiable or measurable external/internal characteristic of the plant, as recorded by the systems and computerized tools mentioned in the present application. Such characteristics can be related to, but are not limited to: height, weight, length of the roots, stem, leaves or branches, length between the nodes, number of nodes, number of fruits, number of flowers or inflorescences, germination, biomass, branching, appearance of adventitious roots, color, leaf shape, woodiness, roots’ weight, transpiration rate, water use efficiency and more. Once those parameters are recorded and collected on plants growing in a small-scaled setting, such as laboratories, they are analyzed by the algorithm of the present invention and result in an output comprising indications and recommendation of how to grow the same plants in a field or in large commercial greenhouses, and how the plants are predicted to perform under similar environmental conditions, previously tested in a small-scale setting.
The term “small-scale settings” generally refers hereinafter to a location or locations in which populations of plants are grown and tested for their performance and tolerance under various environmental conditions by the system and algorithm of the present invention. This may include for example laboratories or small greenhouses in academic or research centers’ facilities. The number of plants grown in those small-scale settings is smaller and the growth period is considered shorter in comparison to the number and period length of plants of the same population, grown in large scale settings, such as open fields. The plants grown in small scale population serve as the population on which experiments and testing are carried out, and their performance parameters are translated by the algorithm of the present invention to predictions and recommendations, concerning future growth of the same plant in large-scale settings. Furthermore, the algorithm of the present application stores data from both small-scale and large scale setting experiments, thus, it is a self-learning algorithm, which is constantly updated. Conducting the experiments in small- scale settings saves time and money, as the growers do not need to wait for the plants’ complete growth cycle and results can be obtained on a relatively small number of plants, when the plants are still relatively young.
The term “large-scale settings” generally refers hereinafter to a location or locations in which populations of plants are grown following the indications and recommendations generated from data obtained from populations of the same plants grown in small-scale settings or from previous large-scale experiments, by the system and algorithm of the present invention. Large-scale settings may include, in a non-limiting way fields, commercial greenhouses etc. The plants grown in those large-scale settings may be grown to a full growth cycle, with the aim of resulting in produce and yield. Furthermore, it is pivotal to clarify that the populations of plants grown in those large-scale settings outnumber the populations of the plants grown in the small-scale settings, and they are grown for commercial purposes.
The term “experiment conditions” generally refers hereinafter to the environmental conditions under which the physiological/phenological/ performance parameters of plant populations (which are not the control group(s)) are evaluated and calculated in small-scale settings. This may be for example, conditions of abiotic stress (such as high concentrations of salt in the soil/water, water deficiency, elevated temperatures etc.). In addition to mere environmental stressful conditions, the term “experiment conditions” may also mean other treatments carried out on the plant populations. For example, seeds can be treated before the small-scale setting experiment with plant hormones or horticultural chemicals, and subsequently be exposed to abiotic stress during the experiment. As a non-limiting example, the present disclosure describes populations of plants tested for their ability to endure salinity stress. The experiment should be carried out in a research lab on a relatively small number of groups for a relatively short period of time (compared to a commercial crop cycle of the same plants in the field). The plant populations may consist of a “control group” (group A) which is grown under optimal conditions (without excessive amounts of salt in the water), a group which is grown with excessive amounts of salt in the water (group B), another group whose seeds were pre-treated with a chemical, and grown with salty water during the experiment (group C) and an additional group, which is a salt-tolerant genotype (group D). Therefore, the term “experiment condition” refers to groups B-D.
The term “algorithm/ a computer implemented non-transitory algorithm/ computer processor” generally refers hereinafter to computer-implementable instructions and specifications designed to execute among other things calculations, data processing and task solving. The disclosed algorithm of the present invention is specifically designed to integrate data obtained from small-scale setting experiments via a phenotypic screening system and a designated software which combined, capture images of the plants and calculates various performance parameters thereof. The algorithm is fed with the above-mentioned data, retrieves previously stored data from past small-scale and large-scale experiments, correlates the environmental conditions of the experiment with the conditions in a large-scale setting, indicates plant populations/individual plants outperforming other plant populations, stores data from a current experiment, and generates a predictive recommendation report based on the data above. The algorithm of the present invention is a self-learning algorithm, which is constantly updated by data obtained from both small-scale and large-scale experiments.
The term “predictive recommendations/indications report” generally refers hereinafter to data obtained from small-scale setting experiments or previous large-scale setting experiments concerning plant performance, physiological/phenological parameters or tolerance, which is analyzed by the algorithm of the present invention. The data are translated to practical, helpful recommendations guiding and instructing growers for example, when to sow seeds in a field, which plant populations/genotypes are more likely to exhibit tolerance to abiotic stress in large- scale setting conditions, what is the percentage of plants within a specific population which would be more likely to survive and perform under stress, if preventive actions of any other crop management nature should be taken, etc. Disclosed are artificial intelligence (Al)-based system and a method for predicting abiotic stress tolerance and performance in tested plant populations. The system provided herein comprises several key components to perform a supervised machine learning algorithm for identifying and/or measuring physiological and phenological parameters within a shorter time frame and on a smaller number of plants than evaluating an entire agricultural crop cycle after determining a significant stress condition by the disclosed AI-assisted system for plant stress mitigation.
The system comprises the following features:
(a) generation of early stage data sets in small-scale settings regarding physiological, phenological, performance or tolerance parameters of various plant populations under defined stress conditions. Laboratory-cultivated plants under unstressed conditions relative to the stressed tested populations serve as a control for these tested parameters.
(b) an algorithm/computerized system that controls and directs the automated plant screening unit. The algorithm collects and evaluates the above data sets acquired by the automated plant screening complex for analyzing and predicting the performance of the current batch of plants in large-scale settings (such as fields) based on the previous data sets. The data sets are constantly updated by data from ongoing screening in the field and in the lab, modifying the selection algorithm.
In various embodiments, the present invention comprises methods for monitoring the physiological/phenological parameters of a selected group of plants grown under environmentally controlled conditions, to screen and select a plurality of plants at early stages of growth for their future phenotypes, analyze the obtained data and interpret it into practical recommendations and indications, for instance, which genotype is more likely to perform better under specific conditions once it is grown in large numbers for commercial purposes. Pluralities of plants could be tested and analyzed by the system and algorithm of the present invention, including in a non-limiting fashion, nursery and wild type plants, as well as transgenic events or mutagenized plants which can be screened by the disclosed system. Additionally, the individual plants are exposed to a controlled environment, and automatically provided with controlled amounts of water, and/or nutrients based upon one or more assay definitions. In other embodiments, the disclosure comprises an automated method for phenotypic screening of a population of plants in small-scale settings using physiological/phenological parameters. The method comprises subjecting the population of plants to an induced abiotic stress in a controlled environment and measuring the physiological/phenological parameters that vary in response to said stress. An unstressed control and a stressed control are present in all the experiments, setting the upper and lower limits for the algorithms for selection.
In a preferred embodiment of the present invention, the screening system and algorithm of the present invention monitors, collects and analyzes data regarding physiological/phenological parameters, performance and tolerance from populations of plants grown in small-scale settings. To better understand the procedures and applications involved in the present invention, the following non-limiting example is disclosed (plant species, number of tested plant groups, growth conditions and parameters may vary depending on the experiment and the growers’ needs and requirements): tomato plants ( Solanum lycopersicum) are grown in a laboratory under controlled conditions, and may include several groups, for example: (a) a group which serves as control in which the plants are grown under optimal/natural/non-stress conditions, (b) one or a plurality of groups grown under different concentration(s) of salt (to induce salinity stress), and (c) one or a plurality of groups which are pre-treated or treated during the experiments with chemicals/biochemical compounds/biological agents/physical treatments etc., and also exposed to said different salt concentrations. The different plant groups are grown and monitored under the surveillance of a phenotypic screening system, which takes measurements of various plants’ physiological/performance parameters. The investigators wish to check if the various screened plant populations exhibit an increase in their biomass (weight) compared to the control group at day 28 from sowing. All the data are collected and analyzed by the algorithm of the present invention. The output of said analysis comprises recommendations and indications for the next crop cycles under those field conditions, predicting how well the same tomato plants would perform in a field, which contains similar concentrations of salts compared with the lab experiment.
In an alternative example, the investigators compare in laboratory settings different genotypes/germplasm of tomato plants grown under the same salt concentration. The data obtained from such small-scale experiment are analyzed and result in a report which specifies which genotype is more tolerant to salinity stress and would be more likely to perform better in a field characterized by having similar salinity conditions.
At the core of this disclosure lays the notion that the disclosed invention simulates the field conditions or large-scale setting on a smaller, lab scale setting and screens the plant populations best suited for that large-scale setting in shorter time.
In yet another preferred embodiment of the present invention, the growers might grow the plants in a large-scale setting (such as a field) having similar environmental conditions as the laboratory experiment and evaluate if the predictive recommendation and indication report is accurate and helpful. The growers report back and give their feedback, so that actual results from large-scale settings are fed into the algorithm of the present invention, ensuring it is constantly updated and capable of retrieving said data in future experiments.
In yet another preferred embodiment of the present invention, the disclosed screening system and algorithm are configured to capture, takes measurements or calculate various physiological/phenological/performance parameters and to present them as an output graphically, visually, statistically, schematically, illustratively or by any other mean acceptable or known in the art.
In yet another preferred embodiment of the present invention, the disclosed screening system and algorithm are configured to compare different plant populations, and additionally carry out comparisons, measurements and calculations within the same population of plants. For instance, when growing and monitoring the plants in a small-scale setting, the screening system and the algorithm may indicate and calculate how many plants from a specific group perform better or are more tolerant to the stress.
In yet another preferred embodiment of the present invention, the disclosed screening system and algorithm are configured to integrate multiple variants in parallel. In other words, the phenotypic screening system is configured to monitor a plurality of physiological/phenological/performance parameters simultaneously on numerous plant populations and extract data concerning all the desired parameters. For instance, the system can monitor at the same time a plurality of plant populations and extract data concerning the plant height, transpiration rate, number of fruits/flowers per plant, plant weight, number of leaves and calculate all those data for each individual plant.
In yet another preferred embodiment of the present invention, the disclosed screening system and algorithm are configured to evaluate different performance parameters in small-scale settings, and determine if those parameters are appropriate and indicative for the designed experiment and the actual, subsequent commercial growth of the plants. The scientific literature is replete with information and data regarding which plant performance parameters should be taken into consideration under abiotic stress. Nevertheless, not all these data or the data obtained from small- scale experiments properly correlate with plant performance in the field. Since the algorithm of the present invention is constantly updated by new results from small-scale settings, real time results and feedback from the large-scale settings and scientific publications, it can filter out parameters which are considered by the scientific literature as reliable or accurate, or parameters considered promising and predictive in past small-scale experiments, in case those parameters did not materialize nor led to satisfactory results in the field. For example, if the literature would indicate plant height as the most important factor for evaluating plants’ tolerance to salinity stress, and in a specific small-scale setting experiment the plants’ height does not significantly vary between tested groups, the algorithm would flag it out in the predictive recommendation report, and highlight other predictive performance parameters, which bear greater significance. Moreover, as the algorithm of the present invention analyzes and integrates a plurality of parameters and data sets obtained from individual plants, and is also a self-learning algorithm, configured to retrieve and assess previous large-scale performance data, the resultant predictive recommendation report is highly accurate.
In yet another preferred embodiment of the present invention, the disclosed algorithm collects and stores data from past small-scale and large-scale setting experiments, and retrieves them for present analyses. The data collected for example, from past large-scale experiments may include in a non-limiting way: yield parameters, produce export quality, characteristics of the salt concentrations in the soil or water and more. In yet another preferred embodiment of the present invention, growers and seed companies are provided with helpful, useful data in the form of predictive recommendations to determine the ability of agricultural crops to mitigate abiotic stress conditions in the field.
In yet another preferred embodiment of the present invention, the analyzed data obtained from the screening system and the algorithm of the present invention may assist growers to better understand when it is productive to initiate a crop management action, as opposed to when it may not be productive to take actions.
In yet another preferred embodiment of the present invention, the system, method and algorithm disclosed herein can save time and costs. As the data is obtained from a relatively small number of plants (compared to the commercial quantities required in a field), and sometime there is no need to wait for the end of the growth cycle in order to obtain the desired data, the present invention can save time and money for growers and seed companies, otherwise spent on large-scale experiments, potentially involving substantial yield loss.
As disclosed herein, the present invention provides an automated AI-based screening system for abiotic stress mitigation in plants which is also predictive of abiotic stress tolerance in large-scale settings.
The following examples are set out, while not limiting the above. The system used in the present set of experiments is designed as one-to-one (1:1) plant- [sensors+controller] unit, i.e., each individual plant is monitored by one unit of sensor, controller and irrigation valves that enable: (i) monitoring water-relation kinetics of each plant and environmental responses throughout the plant’s life cycle with high spatiotemporal resolution, (ii) creating a truly randomized experimental design due to multiple independent treatment scenarios for every plant, and (iii) reducing artificial ambient perturbations due to the immobility of the plants or other objects.
The functional phenotyping system Plantarray 3.0 platform of Plant-Ditech (www.plant- ditech.com/products/plantarray) is an exemplary system which may be used for monitoring plants’ performance during the entire experimental period (in small-scale settings) by controlling irrigation schedule and water quantities. This platform enables performing high-throughput physiological functional phenotyping by continuously, simultaneously and accurately measuring the momentary-varying water flux in the soil-plant atmosphere for each plant in the array. The Plantarray phenotypic system is based on the linear correlation between stomatal conductance and transpiration to plant productivity (i.e. CO2 assimilation), which indicates the plant performance in high correlation to yield results. The system includes up to 72 units of highly sensitive, temperature-compensated load cells that are used as weighing lysimeters. Each unit is connected to a personalized controller, which measures the pot weight 24/7, collects the data and controls the irrigation of each plant separately. An independent controller for each pot enables tight feedback irrigation system, based on the plant’s transpiration rate. Each controller unit is connected to its neighboring unit for serial data collection and all data are loaded to a server. A pot with three plants is placed on each load cell. The data can be analyzed by a designated software, such as SPACanalytics by Plant-Ditech (www.plant-ditech.com/products/spac-analytics), an online cloud- based software that enables viewing and graphically and statistically analyzing the real-time data collected from the Plantarray system. The estimated plant’s weight at the beginning of the experiment is calculated as the difference between the total system weight and the sum of the tare weight of pot + drainage container, weight of soil at pot capacity, and weight of water in the drainage container at the end of the free drainage. The plant’s weight at the end of a growth period (calculated plant’s weight) is calculated as the sum of the initial plant’s weight and the multiplication of the cumulative transpiration during the period by the water use efficiency (WUE). The latter, determined as the ratio between the daily weight gain and the daily transpiration during that day, is automatically calculated on a daily basis by the SPAC-analytics software.
As an alternative to the above plant’s parameters, other plant data such as optical data, reflectance data, x-ray data, thermal emission, audio data (e.g., ultrasonic data, etc.), haptic data, chemical data (e.g., chemical composition of the plant), electric data (e.g., resistivity, voltage open circuit, inductance, electrical noise, conductance, etc.), responsiveness or response to an applied stimulus (e.g., incident light, acoustic noise, haptic stimulus, electric stimulus), thermal data, could form a data set in the system. The collected signals can be within the range of human detection, but can alternatively or additionally be determined based on signals outside the range of human detection. For example, measurements can be taken using or recording an audio frequency outside the aural frequency range or a light frequency outside the visual spectrum. The plant data can additionally or alternatively be chemical characteristics (e.g., chemical composition, concentration of a given chemical, etc.), visual characteristics, electrical characteristics, or any other suitable characteristic of the plant. The plant data are preferably collected from the entirety of the plant body, but can alternatively be collected from a portion (e.g., less than the entirety) thereof, such as a leaf, a flower or a fruit. The plant data are preferably non-destructively collected from the plants, but can alternatively be destructively collected from the plant. The plant data are preferably collected over multiple sessions, spaced over a period of time, such that the plant characteristics are tracked across time. However, the plant data can be collected at any other suitable frequency, for instance twice a day or 3 times a week.
The tested plant populations or any part thereof (such as seeds) can be treated during or before the small-scale settings experiments with various chemicals or biochemical compounds, such as plant hormones, lipids, peptides, proteins PEG, reactive oxygen species etc., biological agents, such as bacteria, fungi, viruses etc., or physical treatments such as irradiation, contact forces, electrical forces and more.
Plant data can be utilized to automatically select plants that express desired phenotypes from the plurality of plants within the plant field. The identified plants are preferably used for successive breeding, wherein the genetic material of the selected plants is preferably used in the creation of the next generation of plants (e.g., by sexual reproduction, genetic vector methods, etc.). Alternatively, the identified plants can be asexually reproduced, such as by taking a cutting and grafting the cutting onto a host plant or by using stem or tuber cuttings. The phenotypic data can additionally or alternatively be used to test individual plant reactions to given treatments or management systems, wherein specific plant phenotypes that are susceptible (or conversely, resistant) to the given treatment can be identified and selected from a field of plants. Individual plant responses to given treatments can additionally be used to make calculated, judicious future planting or cropping decisions.
The selected plants manifesting improved traits in small-scale setting experiments, such as stress tolerance, as detected by the disclosed system and algorithm, can be subsequently recommended to growers to be planted in environments having predicted parameters similar to the microclimate in which the tested plants were grown and evaluated. The selected plants can alternatively be subsequently recommended to growers to be planted with management systems similar to the plant management system and/or plans used for the tested plant population in said small-scale settings. The plants to be screened in the system are preferably agricultural crops, but can alternatively be any other suitable plants. The tested populations could comprise proprietary/ non-proprietary gene pool materials, recombinant DNA products, plants grown from primed, coated or embedded seeds, products of mutagenesis or any other genetic or molecular manipulations.
EXAMPLE 1
The system 100 of the present disclosure is depicted in Fig. 1. A grower/seed company is interested in evaluating and predicting the ability of a specific variety/cultivar/genotype/germplasm of plants to grow and perform in a place, where the soil is known to contain excessive amounts of salt. Seeds of these plants are sown in a research laboratory, in a controlled environment mimicking the conditions of said field. The plants grown in the lab comprise two groups: a control group (plants growing under natural and optimal conditions without stress) and a stress group (plants exposed to high concentrations of salt). When the plants grow and are at a proper stage according to the experimental design, the plant populations 101 are continuously monitored by a monitoring platform 102 (for instance, the phenotyping Plantarray 3.0 platform) for a defined period of time. Each plant is weighed 100-150 times within 24 hours for 10 days, and the amounts of water used for irrigation and irrigation timing are strictly recorded and monitored. The disclosed system 100 collects data regarding phenotypical/phenological parameters from the plant populations (both control and stress groups) and generates corresponding databases 103. After collecting the phenotypical/phenological data sets, the system 100 utilizes a graphical user interface 104 (for example, the cloud-based SPACanalytics software) to generate graphs, calculations and statistical analyses pertinent to the performance of the plant population 101 under salinity stress. Subsequently, the system 100 employs a computer processor 105 (the algorithm of the disclosed application) to: (a) store all the data sets collected from the current experiment in the system’s databases 103; and (b) integrate the collected data sets and convert the graphical and statistical information generated by the graphical user interface 104 into a predictive recommendation report for growers 107, provided the system 100 detects individual plants which perform better under the defined salinity stress. If such tolerant individual plants are detected 106 within the entire plant populations 101, the disclosed system 100 issues a predictive recommendation report 107. In addition to the databases 103 collected from the current small-scale experiment, the computer processor 105 is configured to also access and utilize other optional databases 108, such as data collected and stored from previous small-scale and large-scale settings experiments performed by the system 100, or known data obtained from publicly available resources, such as publications and scientific literature. The predictive recommendation report 107 can comprise for instance, predictions of the percentage of plants from the specific tested variety/cultivar/genotype which would grow and perform well under similar field conditions, recommendations to cross individual salt-tolerant plants with sensitive plants or designing management plans which would assist in the technical maintenance of the plants, should they encounter similar stress conditions in the field.
EXAMPLE 2
The method 200 of the present disclosure is depicted in Fig. 2. The disclosure commences when the need to predict plants’ tolerance to abiotic stress in the field arises. The goal of the disclosed method is to screen relatively small populations of plants in a relatively short period of time and detect individual plants which are more resilient to stress or exhibit enhanced physiological/phenological parameters. Initially, plants of a specific and defined plant populations (of various varieties/cultivars/species/genotypes/germplasm etc.) are grown in a small-scale setting, such as a research laboratory under controlled conditions 201. The plants populations are subjected to control (optimal, stress-free conditions) and to stressful conditions, the likes of which, the same plant populations will experience when commercially grown in large-scale settings, such as fields. In an optional embodiment of this disclosure, the seeds of these plants can be exposed to different treatments prior to sowing, to assess the effect of treatment on stress tolerance. A special phenotyping platform continuously monitors the plants and takes measurements thereof (such as weight, height and transpiration rate) 202, and creates data sets for each plant population 203 (control and stress). Subsequently, a designated software is utilized to graphically and statistically analyze the data obtained from the experiments 204, and then, the special algorithm of the present invention is employed for the interpretation of the data sets 205, by integrating them, analyzing them and selecting the plant populations/individual plants which are likely to better perform and endure abiotic stress in a large-scale setting. The algorithm of the disclosed method is also designed to store the data sets from the current small-scale experiment in the system for future purposes (for example, if a grower will be interested to grow the same type of plants under the tested microclimate, he/she could rely on the current results, predicting the plants’ ability to survive such conditions) 206. Additionally, the algorithm of the present invention is configured to store past data from small-scale setting experiments, large-scale setting experiments and scientific publications, and retrieving them if necessary. More particularly, the disclosed method comprises a step of issuing a predictive recommendation report 207 for the growers, containing all the necessary analyzed data collected during the experiments, and especially single out individual plants/plant populations, whose performance under stress is improved compared to other plants. By using the indications and recommendation of said report, growers would be able to manage their crop-growing more wisely and to select specific plants, which would perform better under abiotic stress in a field, thus resulting in increased yield.
EXAMPLE 3
An example of a potential small-scale setting experiment and its indications follows: seeds of bell pepper ( Capsicum annuum variety “Liad” [10565]) were obtained from Efal-Agri, Israel. For germination, the seeds were sown in a tray with 10-mL cones filled with commercial growing medium (Matza Gan, Shaham, Givat-Ada, Israel), composed of 55% peat (w/w), 20% tuff (w/w) and 25% puffed coconut coir fiber (w/w). The trays were well irrigated and kept in the same greenhouse (on side-tables) where the small-scale experiment was performed. When the seedlings were 4 weeks old, the growing medium was carefully washed off (to avoid damage to the roots) and the seedlings were immediately transferred to 25 L pots filled with 25 -L pots filled with Bental 11 potting soil by Tuff Marom Golan. The volumetric water content (VWC) of the freely drained substrate, noted as pot capacity, was 25 L.
The Physiological Phenotyping Platform - The experiment was conducted in Aug-Sept 2019 in a commercial-like greenhouse (defined in the present disclosure as a “small-scale setting”) located at the Faculty of Agriculture, Food and Environment in Rehovot, Israel. The greenhouse temperature was controlled using fans that blew air through a moist mattress, keeping it below 30°C. The temperature and relative humidity (RH) were 30-37°C and 2-3.5%, respectively. The plants were grown under natural light (midday maximum of 1100 pmol s-1 m-2), representative values for natural conditions during the summer in the central part of Israel, including Rehovot. The temperature, RH, photosynthetically active radiation, barometric pressure and vapor pressure deficit in the greenhouse were continuously monitored by the meteorological station.
The pepper plants were monitored by the Plantarray 3.0 platform, which continuously took measurements of each individual plant. One of the physiological parameters that was being evaluated was the effect of salinity stress on the plant net weight, as depicted in Fig. 3. The plants were divided to the five following populations to be screened:
A) Unstressed control - plants growing under optimal/natural conditions, without exposure to any stress; and the following groups which comprise the “experiment conditions”: B) Stressed control - plants exposed to salinity stress (EC=4.7 - 3000 mg/L);
C) Stressed test population 1 - plants exposed to the above salinity stress, wherein the seeds of these plants were treated prior to the experiment with a plant hormone solution (comprising among others, kinetin, indole acetic acid and gibberellic acid);
D) Stressed test population 5 - plants exposed to the above salinity stress, wherein the seeds of these plants were treated prior to the experiment with Hemoglobin;
E) Stressed test population 8 - plants exposed to the above salinity stress, wherein the seeds of these plants were treated prior to the experiment with PEG 6000.
The results as graphically shown in Fig. 3., present differences in the means of plant net weight over a period of eight successive days (4 Sep 2019-11 Sep 2019). The experiment was repeated 4 times, each repetition comprised a pot of 25 L, each pot containing 3 plants. Pronounced differences in the plants’ weight are clearly and expectedly seen between group A (unstressed control) and all the other screened populations, which suffered the salinity stress. In addition, it is apparent that among all the groups that were exposed to salinity stress, group D (stressed test population 5) manages to mitigate stress better than the other groups over time. Raw data was obtained using the SPAC-analytics software.
Hence, based on this data combined with data from previous large-scale setting experiments conducted on pepper under similar environmental conditions, the algorithm will issue a predictive recommendation report to growers, indicating which seeds, when treated with a specific treatment, can grow better in salt-enriched soil, or which individual resistant plants from a certain population can be crossed with less tolerant varieties/cultivars in order to generate more salt resistant pepper plants. In addition, the algorithm of the present invention is configured to store the data collected from this present small-scale experiment and to retrieve it for future analyses, if necessary. EXAMPLE 4
An example of a comparison between plants’ performance in a small-scale setting and in a large- scale setting experiments as indicated by the predictive recommendation report of the present invention follows: seeds of pepper (Capsicum annuum var. Botaros) were obtained from Zeraim Gedera-Syngenta, Israel. The small-scale setting experiment was conducted in a greenhouse at the Faculty of Agriculture in Rehovot, Israel between 21-07-2020 to 23-08-2020. The small-scale experiment comprised 5 populations of chemically treated plants (“T3 ec 4”; “T5 ec 4”; “T6 ec 4”; “T7 ec 4”; and “T8 ec 4”) and two control populations - “Ctrl ec 1” (unstressed and untreated) and “Ctrl ec 4” (salt-stressed and untreated).
For germination, the seeds were sown in a tray with 10-mL cones filled with commercial growing medium (Matza Gan, Shaham, Givat-Ada, Israel), composed of (w/w) 55% peat, 20% tuff and 25% puffed coconut coir fiber. The trays were well irrigated and kept in the same greenhouse (on side-tables) where the experiment was performed. When the seedlings were 4 weeks old, the growing medium was carefully washed off (to avoid root damage) the seedling roots and the seedlings were immediately transferred to 4-L pots filled with 20/30 sand (Negev Industrial Minerals Ltd., Israel). The numbers 20/30 refer to the upper and lower size of the mesh screen through which the sand was passed (20 = 20 squares across one linear inch of screen).
The functional phenotyping system Plantarray 3.0 platform (Plant-Ditech) was used to monitor the plants' performance during the entire experimental period by controlling the schedule and quantity of irrigation. This platform, which enables performing high-throughput physiological functional phenotyping, includes 72 units of highly sensitive, temperature-compensated load cells that are used as weighing lysimeters. Each unit is connected to its personalized controller, which collects the data and controls the irrigation of each plant separately. An independent controller for each pot enables tight feedback irrigation, based on the plant’s transpiration rate. Each controller unit is connected to its neighbor for serial data collection and loading to a server. A pot with a single plant is placed on each load cell. The data were analyzed by SPAC-analytics (Plant-Ditech), a designated online web-based software that enables viewing and analyzing the real-time data. The data obtained by the Plantarray 3.0 system indicated that one group of plants (referred to herein as “T5 ec 4”) which was subjected to salinity stress exhibited better physiological parameters than its respective control group (“ctrl ec 4”). In contrast, the plant group named herein “T7 ec 4” showed negative growth and development trends compared with the control group. The following tables and figures show several parameters tested and measured by the disclosed system. Table 1 shows cumulative transpiration of the different plant groups, with “T5 ec 4” exhibiting higher values than the control group, and “T7 ec 4” exhibiting lower values compared to control. Reference is also made to Fig. 4A graphically depicting differences in cumulative transpiration in control groups “ctrl ec 1” and “ctrl ec 4” compared to “T5 ec 4”, and to Fig. 4B graphically depicting differences in cumulative transpiration in control groups “ctrl ec 1” and Ctrl ec 4 compared to “T7 ec 4”.
Table 1 - Cumulative transpiration data in small-scale setting experiments
Figure imgf000026_0001
An additional parameter evaluated by the system disclosed in the present invention is the number of flowers, a parameter known to be negatively affected by salinity stress. As demonstrated in Table 2 and graphically depicted in Fig. 5A and 5B, the pepper plants in group “T5 ec 4” had more flowers than its respective control group (37%), whereas group “T7 ec 4”, exhibited a reduction in the number of flowers compared to control (by 13%).
Table 2 - Data on number of flowers in small-scale setting experiments
Figure imgf000026_0002
The dry weight of the whole plant was also monitored and calculated by the system disclosed herein, as part of the small-scale experiment. As demonstrated in Table 3 and graphically depicted in Fig. 6A and 6B, the pepper plants in group “T5 ec 4” exhibited higher dry weight values than control group (19%), whereas group “T7 ec 4”, exhibited a lower dry weight compared to control (by 20%).
Table 3 - Data on whole plant dry weight in small-scale setting experiments
Figure imgf000027_0001
In addition to the dry weight of the entire plant, the dry weight of the roots of t le pepper plants of the different groups was monitored and calculated. Table 4 demonstrates the differences between control group (“Ctrl ec 4”) to tested groups “T5 ec 4” and “T7 ec 4”. The roots of group “T5 ec 4” exhibit an increase of 13% in weight compared to control group, whereas the roots of group “T7 ec 4” weigh less than the roots of control group (by 22%). Reference is also made to Fig. 7A graphically depicting differences in roots’ dry weight between control group and “T5 ec 4”. Fig. 7B graphically depicts differences in roots’ dry weight between control group and “T7 ec 4”.
Table 4 - Data on roots’ dry weight in small-scale setting experiments
Figure imgf000027_0002
After obtaining all the data above, the algorithm of the present invention is deployed. The algorithm is designed to integrate all the data concerning the different plant populations’ performance under said salinity stress, analyze the data and generates a predictive recommendation report for potential growers. The algorithm also retrieves data from previous laboratory and field experiments, as well as from scientific publications, and integrate said data with the data of the current experiment. The report in this case indicates that once grown in large-scale settings (such as fields for commercial purposes), with similar environmental conditions to those of the small- scale setting experiment, plants of the “T5 ec 4” group would be likely to perform better and to result in increased yield despite the abiotic stress. In addition, the report would indicate that plants of the “T7 ec 4” group would result in decreased performance, and therefore, it would be disadvantageous and ill-advised for growers to plant them in this specific environment.
To test if the predictive recommendation report is accurate and reliable, the inventors grew the above-mentioned plant groups in a large-scale setting, which was conducted at the R&D Center of "Mop Yair”, Moshav Hazeva, in the arid Arava region, Israel. Plants were transplanted in 4.08.2020, 3300 plants per 1 dunam. Plants were harvested for measurements and evaluations on the following dates: 22.11.2020, 16.12.2020, and 20.12.2020. In this case, the plants were grown for a longer period of time compared to the small-scale setting experiment.
The experiment’s matrix comprised: the same 5 populations ( “T-3”; “T-5; “T-6”; “T-7”; and “T- 8”) monitored and screened in the small-scale experiment and a control group (referred to herein as “ctrl”), which is also stressed by the environmental conditions of the field, but untreated. The populations were planted in randomized blocks - 4 repetitions, each repetition included 35 plants, where the middle 20 plant are the actual repetition.
The plants transplanted in 3 double rows (each 1.5 m width) without side effect influence. The results obtained from the large-scale setting experiment clearly show that the predictive recommendation report generated by the algorithm of the present invention was accurate. Similar to the small-scale experiment, group “T-5” exhibited better performance than the control population, and group “T-7” was characterized by reduced performance values. The following tables and Figs. 8-9 elaborate the differences between the different plant groups. Table 5 depicts total yield values in kilograms (the total sum of all repetitions) collected till 27.12.2020. It is evident that group “T-5” exhibits higher values than control group (16% increase), whereas group “T-7” exhibits lower yield values compared to control group (9% decrease). Reference is also made to Fig. 8 graphically displaying the differences in total yield between groups “T-5”, “T-7” and control for large-scale settings.
Table 5 - Data on total yield in large-scale setting experiments
Figure imgf000029_0001
An additional yield-related parameter calculated during the large-scale setting experiment was the number of fruits (the total sum of all repetitions). The relevant results are presented in table 6 and Fig. 9. Table 6 shows that there was an increase of 14% in fruit number for the “T-5” group and a 11% decrease in fruit number for “T-7” group, compared to control group. Fig. 9 graphically depicts total fruit numbers in control group, “T-5” and “T-7” groups in a large-scale setting experiment.
Table 6 - Data on total fruit number in large-scale setting experiments
Figure imgf000030_0001
All the yield-related data obtained from these experiments is stored in the algorithm of the present invention and shall be considered as a further data point for the next iteration of recommendations for this variety of Capsicum for subsequent crop cycles under comparable environmental conditions.

Claims

1. An artificial intelligence-based method for screening a plurality of plant populations and predicting plants’ performance and yield under abiotic stress under field conditions, comprising the steps of: a. sowing seeds of said plant populations in controlled small-scale settings under control and experiment conditions; b. monitoring n combinations of predetermined physiological or phenological parameters of each plant of said plurality of plant populations under said control and experiment conditions over a predetermined period of time; c. obtaining n data sets comprising values for said plurality of plant populations based on said predetermined physiological or phenological parameters; d. analyzing said data sets graphically and statistically using a computer implemented non-transitory software medium having instructions to obtain analyzed data sets; and e. employing a computer implemented non-transitory algorithm configured to interpret said analyzed data sets, comprising steps of: i. storing said analyzed data sets in said system; ii. optionally retrieving previous stored analyzed data sets in said system; iii. assessing said analyzed data sets and said previously stored data sets characteristic of said each parameter; and iv. issuing a predictive recommendation report comprising the treatment combinations for field crops of said same plant populations growing in large- scale settings, wherein the environmental conditions in said large-scale settings resemble the environmental conditions characteristic of said small-scale settings.
2. The method of claim 1, wherein said plant populations are agricultural or horticultural plants, selected from a group consisting of monocotyledonous or dicotyledonous plants.
3. The method of claim 1, wherein said plant population are selected from a group consisting of wild type species, cultivars, varieties, genotypes, genetically-modified plants, grafted plants, plants grown from primed, coated or embedded seeds and any combination thereof.
4. The method of claim 1 , wherein said abiotic stress is selected from a group consisting of heat, cold, drought, salinity, osmotic stress, exposure to pollutants, toxins or hazardous chemicals, and physical injuries or wounding and any combination thereof.
5. The method of claim 1, wherein said predetermined physiological or phenological parameters are selected from a group consisting of plant height, plant weight, length of the roots, length of the stem, length of the leaves, length of the branches, length between the nodes, number of nodes, number of fruits, number of flowers or inflorescences, fruit weight, root weight, germination ability, biomass, transpiration rate, water use efficiency, branching, appearance of adventitious roots, color, leaf shape, woodiness, optical data, reflectance data, x-ray data, thermal emission, audio data, ultrasonic data, haptic data, chemical data, electric data, responsiveness or response to an applied stimulus and any combination thereof.
6. The method of claim 5, wherein said electric data are selected from a group consisting of resistivity, voltage open circuit, inductance, electrical noise, conductance and any combination thereof.
7. The method of claim 1, wherein said experiment conditions are selected from a group consisting of: exposing said plant population to abiotic stress, treating said plant populations during the experiment conducted in said small-scale settings, treating said plant populations before the experiment conducted in said small-scale settings, treating the seeds of said plant population before the experiment conducted in said small-scale settings, and any combination thereof.
8. The method of claim 7, wherein said treating is a treatment selected from a group consisting of: exposing said plant populations or seeds thereof to chemicals, exposing said plant populations or seeds thereof to biological agents, treating said plant populations or seeds thereof with physical forces, and any combination thereof.
9. The method of claim 1, wherein said plant populations grown in said small-scale settings are grown in lower numbers and for a shorter period of time compared to said plant populations grown in said large-scale settings.
10. The method of claim 1, wherein said phenotyping system is configured to: (i) calculate linear correlations between stomatal conductance and transpiration to plant productivity; (ii) weigh said plants various times a day; and (iii) control irrigation schedules and water quantities.
11. The method of claim 1 , wherein said computer implemented non transitory software medium is a cloud-based software configured to analyze data collected from said phenotyping system.
12. The method of claim 1, wherein said predictive recommendation report comprises recommendations selected from a group consisting of: prediction of the performance of said plant population in large-scale settings under specific abiotic stress, prediction of the percentages of plants from said plant populations expected to survive said abiotic stress or perform better under said abiotic stress in large-scale settings, defining subpopulations of stress-tolerant plants to be further used for crossing, breeding and cultivation purposes, implementing management systems or protocols, taking cautionary actions and any combination thereof.
13. An artificial intelligence-based system for screening plant populations in small-scale settings and for predicting abiotic stress tolerance in said plant populations in large-scale settings, comprising: a. a phenotyping system; b. a computer implemented non transitory software medium; and c. a computer implemented non transitory algorithm; wherein said phenotyping system is configured to (i) monitor said plant populations under control and experiment conditions; and (ii) generate data sets, said computer implemented non transitory software medium is configured to (i) graphically and statistically analyze said data sets and (ii) result in analyzed data sets, and said computer implemented non transitory algorithm is configured to (i) interpret said analyzed data sets and (ii) generate a predictive recommendation report.
14. The system of claim 13, said plant populations are agricultural or horticultural plants, selected from a group consisting of monocotyledonous or dicotyledonous plants.
15. The system of claim 13, wherein said plant population are selected from a group consisting of wild type species, cultivars, varieties, genotypes, genetically-modified plants, grafted plants, plants grown from primed, coated or embedded seeds and any combination thereof.
16. The system of claim 13, wherein said abiotic stress is selected from a group consisting of heat, cold, drought, salinity, osmotic stress, exposure to pollutants, toxins or hazardous chemicals, and physical injuries or wounding and any combination thereof.
17. The system of claim 13, wherein said experiment conditions are selected from a group consisting of: exposing said plant population to abiotic stress, treating said plant populations during the experiment conducted in said small-scale settings, treating said plant populations before the experiment conducted in said small-scale settings, treating the seeds of said plant population before the experiment conducted in said small-scale settings, and any combination thereof.
18. The system of claim 13, wherein said phenotyping system is configured to (i) calculate linear correlations between stomatal conductance and transpiration to plant productivity, (ii) weight said plants various times a day and (iii) control irrigation schedules and water quantities.
19. The system of claim 13, wherein said computer implemented non transitory software medium is a cloud-based software configured to analyze data collected from said phenotyping system.
20. The system of claim 13, wherein said predictive recommendation report comprises recommendations selected from a group consisting of: prediction of the performance of said plant population in large-scale settings under specific abiotic stress, prediction of the percentages of plants from said plant populations expected to survive said abiotic stress or perform better under said abiotic stress in large-scale settings, defining subpopulations of stress-tolerant plants to be further used for crossing, breeding and cultivation purposes, implementing management systems or protocols, taking cautionary actions and any combination thereof.
PCT/IL2021/050094 2020-01-28 2021-01-27 Screening system for abiotic stress mitigation in plants WO2021152585A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US202062966569P 2020-01-28 2020-01-28
US62/966,569 2020-01-28

Publications (1)

Publication Number Publication Date
WO2021152585A1 true WO2021152585A1 (en) 2021-08-05

Family

ID=77078655

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IL2021/050094 WO2021152585A1 (en) 2020-01-28 2021-01-27 Screening system for abiotic stress mitigation in plants

Country Status (1)

Country Link
WO (1) WO2021152585A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115792185A (en) * 2022-12-05 2023-03-14 山东大学 Method for detecting soil retention effect of foreign species based on stomatal conductance parameters

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3236741A1 (en) * 2014-12-22 2017-11-01 Agenzia Nazionale Per Le Nuove Tecnologie, L'Energia E Lo Sviluppo Economico Sostenibile (ENEA) A microcosm for raising plants under biotic and abiotic conditioning

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3236741A1 (en) * 2014-12-22 2017-11-01 Agenzia Nazionale Per Le Nuove Tecnologie, L'Energia E Lo Sviluppo Economico Sostenibile (ENEA) A microcosm for raising plants under biotic and abiotic conditioning

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
ALHNAITY, B. ET AL.: "Using deep learning to predict plant growth and yield in greenhouse environments", INTERNATIONAL SYMPOSIUM ON ADVANCED TECHNOLOGIES AND MANAGEMENT FOR INNOVATIVE GREENHOUSES: GREENSYS2019 1296, 1 June 2019 (2019-06-01), pages 425 - 432, XP081387898, Retrieved from the Internet <URL:https://arxiv.org/ftp/arxiv/papers/1907/1907.00624.pdf> *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115792185A (en) * 2022-12-05 2023-03-14 山东大学 Method for detecting soil retention effect of foreign species based on stomatal conductance parameters

Similar Documents

Publication Publication Date Title
Gosa et al. Quantitative and comparative analysis of whole-plant performance for functional physiological traits phenotyping: new tools to support pre-breeding and plant stress physiology studies
Nemeskéri et al. Physiological responses of selected vegetable crop species to water stress
Poudyal et al. Phenotyping from lab to field–tomato lines screened for heat stress using Fv/Fm maintain high fruit yield during thermal stress in the field
Burgess et al. Exploring relationships between canopy architecture, light distribution, and photosynthesis in contrasting rice genotypes using 3D canopy reconstruction
Constable et al. The yield potential of cotton (Gossypium hirsutum L.)
Halperin et al. High‐throughput physiological phenotyping and screening system for the characterization of plant–environment interactions
Gorim et al. Evaluation of wild lentil species as genetic resources to improve drought tolerance in cultivated lentil
Wijewardana et al. Screening corn hybrids for cold tolerance using morphological traits for early‐season seeding
Morison et al. Improving water use in crop production
Boote et al. Improving the CROPGRO-tomato model for predicting growth and yield response to temperature
EP3897103A1 (en) A method of finding a target environment suitable for growth of a plant variety
Kuan-Hung et al. Applying Dickson quality index, chlorophyll fluorescence, and leaf area index for assessing plant quality of Pentas lanceolata
Gorim et al. Root traits, nodulation and root distribution in soil for five wild lentil species and Lens culinaris (Medik.) grown under well-watered conditions
Belachew et al. Association of shoot and root responses to water deficit in young faba bean (Vicia faba L.) plants
Kim et al. Application of maximum quantum yield, a parameter of chlorophyll fluorescence, for early determination of bacterial wilt in tomato seedlings
Zhang et al. Variation of photosynthetic induction in major horticultural crops is mostly driven by differences in stomatal traits
Alsajri et al. Morpho-physiological, yield, and transgenerational seed germination responses of soybean to temperature
Fleisher et al. Effects of CO2 and temperature on crops: Lessons from SPAR growth chambers
Polania et al. Contrasting Phaseolus crop water use patterns and stomatal dynamics in response to terminal drought
Tshikunde et al. Leaf gas exchange and water-use efficiency of dry-land wheat genotypes under water stressed and non-stressed conditions
WO2021152585A1 (en) Screening system for abiotic stress mitigation in plants
White et al. Crop response to climate: ecophysiological models
Pandey et al. Leaf potassium status for drought tolerance: the hunt for promising sesame (Sesamum indicum L.) accessions
Rao et al. Phenotyping horticultural crops for abiotic stress tolerance
Hu et al. A new method to estimate vegetable seedling vigor, piloted with tomato, for use in grafting and other contexts

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21747088

Country of ref document: EP

Kind code of ref document: A1

DPE1 Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101)
NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21747088

Country of ref document: EP

Kind code of ref document: A1